“DeepMASC: Automated Mask Creation and Class Selection in Single Particle Analysis Empowered by Deep Protein Probability Measures of Cryo-EM Maps”

_

Date and Time

4:00 PM – 5:30 PM Monday, June 16^th, 2025

Location

Hybrid (Zoom and KEK SBRC building)

---

Speakers:

Mr. Han Zhu, M.S.

Department of Computer Science, Purdue University, US.

Abstract

We have developed DeepMASC (Deep Masking and Auto-Selection for Classifications), a novel approach that eliminates human intervention from two critical steps in the cryo-EM single particle analysis (SPA) computational workflow: 3D class selection and 3D reference mask generation. Our method leverages deep learning-based protein probability measures to automatically identify optimal structural classes and generate appropriate masks for further refinement. DeepMASC dramatically accelerates the SPA workflow by eliminating time-consuming manual interventions that often create significant delays. Traditional workflows require expert decisions at critical junctures, creating bottlenecks that can extend processing time from hours to days. By automating these decision points, DeepMASC enables continuous, uninterrupted processing without sacrificing quality.
To address broader computational challenges in cryo-EM analysis, we also have developed the Kihara Lab EMSuite Server, a free and comprehensive web platform offering 13 algorithms for cryo-EM analysis. The server spans the full spectrum of needs, from secondary structure detection at medium resolution to atomic modeling at high resolution for proteins, nucleic acids, and their complexes. This unified interface makes sophisticated cryo-EM analysis accessible to structural biologists regardless of computational expertise, democratizing access to cutting-edge methods without requiring specialized hardware or software installation.

Profile

2018-2022              B.S., Department of Computer Science, Purdue University, US.

2022-2023              M.S., Department of Computer Science, Purdue University, US.

2023-Present          Ph.D. program, Department of Computer Science, Purdue University, US

Research Focus

Computational Structural Biology, ML-Based Cryo-EM Modeling and Analysis

---

Click here to apply

“Tracking molecular motions in proteins using conventional and advanced X-ray crystallography techniques”

_

Date and Time

4:00 PM – Wednesday, May 21, 2025

Location

Hybrid (Zoom and KEK CryoEM building)

---

Speakers:

Swagatha Ghosh

Designated Assistant Professor, Department of Applied Physics, Nagoya

University (Japan)

Abstract

X-ray crystallography remains one of the most widely used and well-established structural biology techniques for obtaining high-resolution, atomic-level structures of macromolecules. While most X-ray crystallography proteins structures till date emerge from single crystals shot at cryogenic temperatures, and automation techniques have immensely increased their utilization, these conventional methods have certain drawbacks. Requirement of large crystals and screening for suitable cryo-protectants could limit structural studies of proteins at physiologically relevant conditions and triggering reaction for real-time dynamical information. Recently, advancements of serial X-ray crystallography techniques with micro-focused beams and faster detectors have allowed structure determination from (sub)micron- sized crystals at room-temperature and deciphering molecular mechanisms triggered with light or chemical stimulus. This technology hence allows generating ‘molecular movies’ by capturing different transient states of a protein’s reaction cycle.
In my talk, I will give an overview of my research on several soluble and membrane protein systems with exciting properties and utilization of various X-ray crystallography techniques for elucidating their molecular mechanisms. Briefly, I will discuss about 1) my on-going project on deciphering photo- physical mechanisms of large spectral shift fluorescent proteins from blue fish, 2) several past projects on membrane proteins associated with bioenergetics in cells and 3) design and development sample-delivery devices for in situ and serial crystallography that could be implemented in any synchrotron radiation facility. The goal of my presentation is to promote a simpler approach for sample preparation and data-collection that could lower the entry barrier for novice users for X-ray crystallography at synchrotrons.

Presenter’s bio

Swagatha Ghosh is a structural biologist with keen interest in macromolecular biophysics of proteins using conventional and advanced X-ray crystallography and UV-VIS spectroscopy. She had completed her PhD in Bangalore, India and postdoctoral research in Gothenburg, Sweden. Using conventional- and serial- crystallography techniques, her research aims at deciphering structural mechanisms in proteins. She has co-founded a deep-tech startup in Sweden aiming to develop sample
delivery techniques for serial crystallography. She moved to Nagoya University (Japan) in August 2022, as Designated Assistant Professor to pursue her research on fluorescent proteins (FP). In addition to her research, she teaches experimental biophysics to undergrad students and is actively involved in building networking platforms for international researchers at Nagoya University.

---

Click here to apply

“The structure, dynamics and function of Adaptive Immune Receptors ”

_

Date and Time

4:00 PM – 5:30 PM Friday, May 9^th, 2025

Location

Hybrid (Zoom and KEK CryoEM building)

---

Speakers:

Daron Standley, PhD

Research Institute for Microbial Diseases (RIMD), Osaka University (Professor)

Abstract

Adaptive immune receptors (AIRs) consist of B cell receptors (BCRs)—also known as antibodies in their soluble form—and T cell receptors (TCRs), which remain anchored to the plasma membrane of T cells. AIRs are unique among human proteins in that they are generated by the combinatorial assembly of genes, leading to an extraordinary diversity of expressed proteins that undergo positive and negative selection throughout our lives. The diversity is so great that any two humans typically share only a few percent of their BCRs or TCRs in a routine blood draw. In spite of this incredible diversity, different individuals respond to vaccines in much the same way within days—a remarkable example of robustness and functional convergence. I will describe one particularly interesting example of convergent responses that, unexpectedly, enhances SARS-CoV-2 infection. I will also show how BCR and TCR sequencing of peripheral blood can be used to diagnose a wide range of diseases, including infections, autoimmunity, and cancer. If time permits, I will also introduce a model of how T cells sense antigens—derived from unbiased molecular dynamics simulations and validated experimentally.

Profile

1998–2003: Schrödinger, Inc. (New York, USA) — Scientific Software Developer
Engaged in the development of scientific software, including flexible docking.

2003–2008: Osaka University, Institute for Protein Research, Protein Data Bank Japan (PDBj) — Senior Researcher
Developed tools for structure-based searching  at PDBj.

2008–2014: Osaka University, Immunology Frontier Research Center (IFReC) — Specially Appointed Associate Professor
Launched the systems immunology lab where I worked on modeling immune responses.

2014–2016: Kyoto University, Institute for Virus Research — Professor
Developed tools for high-throughput BCR and TCR modeling.

2014–Present: Osaka University, Research Institute for Microbial Diseases, Division of Genome Informatics — Professor
I worked in disease diagnosis using adaptive immune receptors.

---

Click here to apply

“SBCloud: A Cloud-based Platform for Cryo-EM Structure Determination and Training from SBGrid”

_

Date and Time

4:00 PM – 5:30 PM Wednesday, April 16^th, 2025

Location

Hybrid (Zoom and KEK CryoEM building)

---

Speakers:

Jason Key, PhD

Associate Director of Technology & Innovation, SBGrid Consortium Lecturer on Biological Chemistry and Molecular Pharmacology,
Harvard Medical School

　　

Abstract

The ever-growing demand for computational power, data storage, and access to cutting-edge scientific software can be a formidable challenge for structural biology research labs. SBGrid, an international research computing consortium based at Harvard Medical School aims to break down technological barriers in structural biology and accelerate discoveries. To support this mission, SBGrid provides an extensive suite of scientific software, expert domain-specific support, computing hardware solutions, and training for scientists investigating molecular mechanisms at the atomic level. Recognizing the increasing computational demands of cryo-EM structure determination, we developed SBCloud—an AWS infrastructure-as-code high-performance computing environment designed for advanced analysis and visualization of structural biology data. We have used SBCloud to host global cryo-EM workshops, fostering collaboration and hands-on training in high-performance computing for structural biologists. We recently offered SBCloud to the SBGrid community on a limited basis, enabling researchers to tackle cryo-EM and Cryo-ET structure determination projects and enabling us to test our platform at scale. In this presentation, I will share insights and lessons learned from our experience in training and supporting the global cryo-EM and Cryo-ET research community and discuss the role of cloud resources in computational structural biology.

　　

Profile

1997 BS Biochemistry
Purdue University, West Lafayette, IN, USA

2004 Ph.D, Time-resolved Laue diffraction of Bacterial sensor proteins
University of Chicago, Chicago IL, USA

2004-2007, Postdoc, Femtosecond laser spectroscopy of sensor proteins
Free University Amsterdam/University of Amsterdam, Netherlands

2007-2012, Postdoc, Inhibition of HIF-2alpha with small molecules
University of Texas Southwestern, TX USA

2012-2014, Structural Biology Computing Specialist, Harvard
2014-current, Lecturer on Biological Chemistry and Molecular Pharmacology
2014-current, Associate DIrector of Technology and Innovation, SBGrid Consortium

---

Click here to apply

“Structural Biology in Industry – Proteros overview”

_

Date and Time

5:00 PM – 6:30 PM Wednesday, January 22^th, 2025

Location

Online (Zoom)

---

Speakers:

Eva Cunha, PhD

Profile:
2006 Master (Bioinformatics), University of Lisbon, Portugal
2013 Ph.D. (Biophysics), Jonhs Hopkins University, USA
2013-2015 Postdoctoral Researcher with Dr. N. Abrescia, CIC-Biogune, Spain
2016-2018 Postdoctoral Researcher with Dr. W. Kuhlbrandt, Max-Planck-Institute for Biophysics, Germany
2018-2021 Researcher, University of Oslo, Norway
2021-2022 Principal Investigator, University of Oslo, Norway
2022-2023 Director for cryoEM, Proteros Biostructures, Germany
2024-Present Department Head and BU co-head for cryoEM, Proteros Biostructures, Germany

Stephan Krapp, PhD

Profile:
1997 Master of Biochemistry, Imperial College, UK / University of Bayreuth, Germany
2002 Ph.D. in Protein Crystallography, Max-Planck-Institute of Biochemistry, Martinsried, Germany (Prof. Dr. Robert Huber)
2003 Research scientist and project manager at Proteros Biostructures, Martinsried, Germany
2010 Department Head of Structural Biology at Proteros Biostructures, Martinsried, Germany
2022 – present: Business Unit Head x-ray crystallography and co-Department head x-ray crystallography at Proteros Biostructures, Martinsried, Germany

Abstract

Proteros offers cutting-edge cryo-electron microscopy (cryoEM) and x-ray crystallography services tailored to accelerate drug discovery by providing high-resolution structural insights into challenging pharmaceutical targets. Our platform integrates state-of-the-art cryoEM and x-ray instrumentation, proprietary workflows, and deep expertise in structure determination to deliver actionable data for small-molecule and biologics programs.

Proteros specializes in resolving protein structures, protein-ligand complexes, and multi-protein assemblies with precision, even for challenging targets like membrane proteins. By enabling atomic resolution insights, our services empower researchers to drive structure-based drug design, optimize lead compounds, and elucidate mechanisms of action. This talk will highlight Proteros’ streamlined processes, success stories, and how our structural biology capabilities bridge the gap between molecular structures and therapeutic breakthroughs. Case studies will include GPCR´s (e.g. GLP1R), ion channels (e.g. Kv.1.3) Kinases (e.g. TAK1, cRAF-14-3-3) and others.

---

Click here to apply

構造解析ソフトウェアRELION (Sjors Scheres, MRC-LMB)を中心としたデータ解析講習会をオンライン開催いたします。

クライオ電子顕微鏡によるタンパク質の単粒子解析に興味をお持ちの方、実際にご自身では解析されたことがない初心者の方を対象にしています。
講習会では、RELIONのチュートリアルに沿って解説および座学を実施する予定です。開催者側でオンライン利用できる解析環境を用意いたします。参加者の方には実際に手を動かして、解析の流れを経験していただきます。

【開催日時】

1日目：2024年11月21日（木）10:00 〜 17:45 終了予定 （18:00 〜 19:00 オンライン懇親会）
2日目：2024年11月22日（金）10:00 〜 17:45 終了予定

【会場】

オンライン開催（Zoom）
講習会にアクセスするためのZoom URLは、参加登録時にお申込みいただきましたメールアドレスにお送りいたします。開催日が近づきましたら改めて個別にご案内をさせていただきます。

【言語】

日本語

【応募〆切 2024年11月8日（金）15:00】

申込数が上限に達しましたので、実習参加は締め切りました。

【募集対象と人数】

本講習会は、クライオ電子顕微鏡解析のご経験がないか、ほとんど無い方を対象としています。なお、学生の方は指導教員に了解を得ていただきますようお願いします。

定員：20名 （解析計算実行環境のリソースに限りがあるため、人数の制限をさせていただいています。予め、 ご了承ください。）

【参加条件】

今回の講習会は、クラウドコンピューター(AWS)上にRelionによる解析環境を開催者側で用意し、そこに参加者の皆様のパソコンからインターネットを通じてアクセスし、解析していただくことになります。従って、高性能のCPUやGPUを掲載したノートパソコンをご使用になる必要はございません。通常、ラボでデータ取りまとめ用などで使われているものでも、十分お使いいただけるものと考えています。Windows、Mac、どちらでも構いません。

民間企業の方のご出席につきましては、KEKのCryoEMコンソーシアム参加企業の方に限定させて頂いております。コンソーシアムへの入会は随時受付けておりますので、ご入会方法の詳細などについては下記の事務局連絡先までお気軽にお問い合わせください。

【参加費】

参加費：無料、その他の補助はありません。

【講師一覧】

重松秀樹：JASRI/SPring-8
安達成彦：筑波大学・生存ダイナミクス研究センター
荒牧慎二：Tietz Video & Image Processing Systems GmbH
川崎政人：KEK・物質構造科学研究所・構造生物学研究センター
守屋俊夫：KEK・物質構造科学研究所・構造生物学研究センター
中村　司：KEK・物質構造科学研究所・構造生物学研究センター
池田聡人：KEK・物質構造科学研究所・構造生物学研究センター

【関連事業】

創薬等先端支援技術基盤プラットフォーム（BINDS）

【主催】

KEK・物質構造科学研究所・構造生物学研究センター
クライオ電顕ネットワーク

【プログラム】

プログラムの詳細は後日公開いたします。
前回のプログラムはこちら

---

＊プログラムの変更があった場合は、随時更新します＊

11月21日（木）

10:00-10:10　はじめに　池田聡人（KEK）
10:10-10:40　座学：Relion5-betaデータ解析〜２次元電顕像から初期３次元構造再構築まで〜
10:40-11:00　講演：AWSデータ解析環境の概要紹介と使い方　中村司（KEK）
11:00-11:20　実習：システムへのログイン、Relion起動
11:20-11:40　実習：プロジェクト作成、ドリフト補正処理開始
11:40-12:00　実習：ドリフト補正結果検証、CTF推定処理開始・結果検証

12:00 13:00　昼食

13:00-13:30　実習：手動粒子ピッキング
13:30-14:10　実習：自動粒子ピッカー訓練処理開始
14:10-14:40　座学：二次元クラス平均化までの基礎理論
14:40-14:55　実習：自動粒子ピッキング処理継続・結果検証
14:55-15:10　実習：二次元クラス平均化処理開始

15:10-15:40　休憩

15:40-16:10　実習：二次元クラス平均化処理結果検証、クラス選別、粒子抽出
16:10-16:30　講演：KEKクライオ電顕施設の紹介と利用方法　川崎政人（KEK）
16:30-16:45　実習：初期三次元構造再構築処理開始
16:45-17:30　招待講演：重松秀樹（JASRI/SPring-8）
17:30-17:45 　実習：初期三次元構造再構築結果と掌性（handedness）検証

18:00-19:00　懇親会（オンライン）

　

11月22日（金）

10:00-10:10　座学：Relion5-betaデータ解析(概要)〜電顕鏡画像の問題点とそれを解消する３次元分類と精密化〜
10:10-10:30　実習：三次元クラス分類処理開始
10:30-11:10　座学：三次元処理の基礎理論
11:10-11:50　実習：三次元クラス分類処理結果検証、クラス選別、粒子抽出
11:50-12:10　実習：三次元精密化処理開始
12:10-13:10　昼食
13:10-13:40　実習：三次元精密化結果検証
13:40-14:10　実習：マスク作成と後処理開始・結果検証
14:10-14:40　実習：CTF精密化処理開始・結果検証、三次元精密化処理開始

14:40-15:10 写真撮影、休憩

15:10-15:30　実習：三次元精密化処理結果検証、ベイズポリッシュ処理開始・結果検証、三次元精密化処理開
15:30-16:00　講演：Beyond Relion5 Tutorial　守屋俊夫（KEK）
16:00-16:10　実習：三次元精密化処理結果検証
16:10-16:25　実習：局所分解能推定処理開始・結果検証
16:25-17:05　講演：ダイナミクス解析と原子モデリング　中村司（KEK）
17:05-17:20　講演：crYOLOと全自動解析処理・結果検証　池田聡人（KEK）
17:20-17:25　終わりに　守屋俊夫（KEK）

---

【 問い合わせ】　

オーガナイザー
KEK・物質構造科学研究所・構造生物学研究センター
池田聡人
E-mail: akihito.ikeda[at]kek.jp

事務局
KEK・物質構造科学研究所・構造生物学研究センター
増田千穂 
電話：029-879-6176
E-mail: cmasuda[at]post.kek.jp